CSIRO has confirmed that a controversial artificial intelligence trial at science magazine Cosmos has been paused.
The revelation came on the same day the National Science Statement was released, which sounded a cautious note on the advance of generative artificial intelligence, warning the technology poses “new risks to trust in institutions and the communication of trusted scientific advice.”
Cosmos is a new addition to the CSIRO Publishing stable. In June, CSIRO – with funding assistance from the Commonwealth and South Australian Governments – announced a one-year deal to take on the publication of Cosmos from embattled former publisher the Royal Institution of Australia (RiAus).
At the time, South Australian Deputy Premier Dr Susan Close described “the importance of timely, accurate and trustworthy reporting on scientific issues… this investment in Cosmos magazine – as well as its journalists, associated publications and educational resources – is an investment against misinformation.”
According to a statement from CSIRO, the Cosmos artificial intelligence (AI) trial occurred when former publisher RiAus applied for a funding grant from the 2023 Meta News Fund (administered by the Walkley Foundation) for “a project which aimed to investigate the opportunities and risks of using artificial intelligence in supporting science communication professionals to develop and disseminate important, factually accurate, science explainers.”
An investigative report from ABC Science exposed the Cosmos project to public scrutiny and attracted criticism and concern from former editors and freelance staff, as well as current contributors who cited a lack of awareness or consultation over the AI roll out.
“At a time when trust in scientific expertise and the media are both declining – the latter more precipitously than the former – rolling out an AI experiment with a lack of transparency is, at best, ignorant, and, at worst, dangerous,” said Jackson Ryan, President of the Science Journalists Association of Australia.
Four days after the original ABC Science report, and on the same date as the release of the National Science Statement, CSIRO announced that Cosmos was pausing the AI project.
“Cosmos has briefly paused the project in order to review the feedback and queries received so far and remains committed to ensuring responsible and ethical AI practices.”
Days earlier, the ABC reported that a CSIRO spokesperson had dismissed concerns and backed the “experimental project” which is designed to investigate the “opportunities and risks of using AI.”
In a statement confirming the pause, Cosmos defended the project, characterising the initiative more as a voyage of scientific inquiry rather than a cost cutting measure.
“We have made the decision to cross the divide and become part of a science experiment to answer important questions about the use of AI in the media. Our aim is to report back on our experiences of using AI in our newsroom and provide evidence-based examples of the risks and opportunities for science communication.”
“Our first learning, with the benefit of hindsight, is that we needed to more clearly communicate the project before testing and publishing trial articles. We could have done better in that respect. We have briefly paused the project while we review the feedback and queries received so far and remain committed to ensuring responsible and ethical AI practices.”
However, former publisher and RiAus Executive Director Will Berryman said the grant application for an AI project had Cosmos’s bottom line in mind.
In comments made to the ABC, Mr Berryman said AI tools like Cosmos’s were needed to push back against misinformation, which was being increasingly AI-generated, and to help publishers produce content at a lower cost, given the decline of advertising revenue that used to keep magazines operating.
“The business model is broken for things like Cosmos,” he said.
Jackson Ryan points out that the revelation of the AI project came just four months after Cosmos made five of its eight staff redundant.
“Both editorial staff and readers should be privy to the reason why an outlet might start using generative AI and where it will do so.”
“There can be no secrecy or subterfuge. That, as we’ve seen time and again, is how you destroy trust,” Mr Ryan said.
The timing of the controversy at Cosmos is particularly embarrassing for CSIRO given the warning on AI sounded by the Federal Government and Australian Chief Scientist Cathy Foley as part of the National Science Statement.
“Advances in generative AI pose new risks to trust in institutions and the communication of trusted scientific advice. This could lead to an increase in plausible sounding, but false, scientific narratives.”
“At their worst, these technologies could see scientific literature being used to advance malicious products, such as harmful drugs, at greater speed and scale than has previously been possible,” the statement says.
